From Silicon to Darude – Sandstorm: Breaking Famous Synthesizer Dsps [video]

It’s great to see some synth chip hacking at CCC this year .. in fact, the synth DSP hacking scene has been great lately .. since the Motorola 56K is EOL’ed, The Usual Suspects folks built an emulator, and its capable now of running the firmware for many popular synths which used that chip:

https://dsp56300.wordpress.com/

This means you can emulate with full accuracy the Access Virus A, B, C, TI / Clavia Nord Lead 2X, 3, Modular / Waldorf Q, Microwave II / Novation Supernova, Nova .. and they’re starting to target others ..

There are plenty of other 56K DSP apps out there - not just synthesizers - so if this proves effective it could mean a lot to certain folks who still have big deployments of these chips, in terms of product life-cycle.

Is it possible (practical) to reverse engineer a synth that uses a custom and undocumented DSP? The Alesis Micron is probably my favorite synth and I have always wanted it in a little laptop groovebox type of package, considered just repackaging mine but the main PCB is not very suited to what I would want. I also tried to just redesign it from the ground up but it has some tricks for avoiding aliasing and are a major part of its overall sound which I can not figure out.

Assuming this is possible, any resources for learning how to go about it? What little I have found has relied heavily on the DSP being well documented.

Apologies if OP video goes into this, internet is not cooperating right now and steaming is not going to happen. Looking forward to watching it, never pass on anything that might offer even a tiny glimpse into reverse engineering the Micron.

It's amazing to see the incredible effort it takes to reverse engineer a synth. While much of the process described in the video went over my head I still found it very interesting to watch. Kudos to the Usual Suspects for not only putting in all that effort but also releasing their findings completely for free.

Side note: does anyone know what the song played at 4:30 is called? It sounds very familiar but I can't remember its name at all.

It's interesting that the Roland DSP does eloquently map to solutions in the z-domain, and allow easy implementation of FIR/IIR filters.

It's not surprising they doubled the clock rate to 88.2kHz to simplify anti-aliasing, but it is curious the choice of the CD audio frequency of 44.1kHz and not the DAT tape frequency of 48kHz.

It's hard not to get nostalgic about the 90s when watching this video.

One of the most incredible talks and demonstrations. Excellent technical work, great team!